In practical applications of non-contact vital sign detection, the noise caused by irregular body movement presents severe interference for accurate detection of vital signs such as respiration and heartbeat signal. Since random body movement is comparable or even stronger than the weak vital sign signal, to some extent it is the main factor limiting the applications of non-contact vital sign sensors.
To reduce body movement, prior art techniques require that the subject or patient being monitored remain stationary and motionless. In many applications such as healthcare, sports, law enforcement, security, it is difficult if not impossible to have the subject being monitored to remain stationary.
Accordingly, what is needed is a method and a system to overcome the aforementioned problems and to recover severely distorted signals during non-contact vital sign detection even when the subject is not stationary.